DESCRIPTION (Applicant's Abstract): Phase I/II clinical trials by the applicants using ricin A chain-containing immunotoxins (RTA-ITs) in over 200 lymphoma patients have established that vascular leak syndrome (VLS) is the dose-limiting side effect. If VLS could be partially or completely reversed without reducing the efficacy of RTA-ITs, larger tumors could be treated and side effects would be minimal. Based on she past experience, the applicant has found that 1) RTA-IT induced VLS occurs in humans but not experimental animals, including mice, rats and monkeys, 2) RTA directly damaged human vascular endothelial cells (ECs) in vitro and this effect could be reversed by the addition of fibronectin (Fn), 3) patients with the most severe VLS had the greatest decreases in levels of their serum Fn. Based on these findings, the applicant now proposes to 1) identify some of the events involved in IT- mediated VLS. In particular, she will focus on ricin A chain (RTA)-based ITs. To this end, she will use the in vitro human-umbilical vein endothelial cell (HUVEC) model, the in vivo syngeneic mouse lung and the SCID mouse/human skin xenograft models to study the effects of ITs on human and mouse vasculature. 2) Investigate the possibility of decreasing VLS in human tissue by the administration of agents which inhibit different steps in a hypothesized pathway leading to VLS. 3) Identify and then mutate/delete VLS-inducing sequences in RTA. The applicant will generate conjugates of "VLS-inducing" candidate peptides from RTA and conjugate them to mouse IgG. She will then study the ability of these conjugates to induce VLS in the three models. If an RTA sequence is identified, it will be deleted from, or mutated in, recombinant RTA. The altered RTA DNA will be expressed, the protein purified, and analyzed for both enzymatic activity and VLS-inducing activity. Taken together, these aims will help to understand and hopefully prevent RTA- IT-mediated VLS in patients.